I thought SY were supposed to be experts at this sort of thing. Perhaps they should have contracted it out to G4S or GCHQ
I am sure they have sufficient expertise to deal with a data source of this size. Size was problematical for the PJ in 2007, but 2007 was when things changed and cheap, off the shelf software became available that handle data pools many thousands of times bigger. Size simply is no longer an issue. My spreadsheet in Open Office laughs at 74,000 items.
Then we hit a second issue. Neither the PJ nor SY were/are familiar with Luz, and that happens to be a magic ingredient.
Then we hit a third issue. The phone traffic covers Portugal, the UK, Germany, Holland, Ireland, and probably others such as Switzerland and Spain. We are not talking small or easy here, whether in PJ days or SY days.
However, the issue raised by Andy Redwood is PAYG phones, so let me give you one example from that. I am sitting on a 2011 Blue Pages guide, supposedly commercial, but with an awful lot of mobile numbers in it. One section is used for 'adults only', and I would guess most of those numbers are PAYG, but identified in my phone book.
I have passed to SY 10 land-line numbers for Luz organisations that a) no longer are in existence and b) do not turn up in a Yellow Pages check. I also passed them 10 mobile numbers that are Luz related and which do not turn up in either a Yellow Pages check or a White Pages check. I simply took these as a sample. I am sitting on more and most households in Luz are sitting on a stash equivalent to mine, and that is thousands of such stashes.
Here is some speculation on my part. The 2014 digs to the east of Luz seem to match up to Euclides Monteiro aka tractorman. If his phone was active in Luz that evening, one has to ask why. The data probably, but not certainly, answers that. The 'correct' pattern would give him the alibi of not being near 5A at the time, even if he was in Luz. The 'incorrect' pattern, should it place him in Luz, would not rule him out, but it would not prove guilt.
The PJ concluded that the phone data could not be boiled down - there had to be suspects first, which could then be analysed. SY seems unable to boil the data down. I'm confident the data can be boiled down much further. As to whether the result is useful, I clearly cannot tell.